Disease Prediction using Machine Learning

Inspiration

Due to the Covid-19 pandemic many hospitals are filled to maximum capacity. There are also many people who are afraid to go to hospitals or doctors due to the pandemic. So we decided to create an easy to use application that would help them diagnose themselves and decide whether they should risk a trip to their doctor.

What it does

Our project is a website that outputs a disease/ailment depending on which symptoms were checked off. It computes the disease using a machine learning model.

How we built it

We built the machine learning model using a dataset that we found online. We used python as our primary programming language. We used pandas to import the training dataset and trained our model using a logistic regression algorithm. The machine learning was done with scikit learn. We then proceeded to test our machine learning algorithm using a test dataset. We then created a development web server using flask and proceeded to create multiple html files as templates to be used for the aforementioned website. We created a checkbox for each symptom. We then imported the chosen symptoms as values corresponding to their column number on the training dataset. We then converted the values into ones and zeros depending on whether the box was checked and ran them through our machine learning algorithm. The machine learning algorithm would return a disease/ailment. We then redirect the user to a different page with the disease/ailment they have and a helpful link with more information about their disease/ailment.

Challenges we ran into

One challenge that we faced was getting the dataset to import properly as it would import with a column of nans. Another challenge was getting the website to output the users choices back to us. A third challenge was getting the prognosis page to redirect properly with the prognosis. Also none of us had much experience in html, css, and machine learning.

What we learned

What we learned from this Hackathon was, using python and machine learning packages such as sci-kit (specifically logistic regression in sci-kit) to train models, using python frameworks such as Flask for web development, and using HTML, CSS, and Jinja2 to develop interactive web pages.

What's next for Disease Prediction using Machine Learning

What's next for Disease Prediction using Machine Learning is adding more symptoms and training the model to predict more diseases.

Accomplishments that we're proud of

Accomplishments that we are proud of are developing a machine learning model that accurately predicts diseases based upon a set of one or more symptoms. We are also proud of successfully developing a website.

Built With

• css
• flask
• html
• jinja
• numpy
• pandas
• python
• replit
• scikit-learn